Using Machine Learning To Help Refugees Succeed

The Stanford Immigration Policy Lab has developed GeoMatch, a machine learning tool designed to optimize the placement of refugees in host countries to enhance their employment prospects. By analyzing individual refugee characteristics—such as country of origin, language skills, gender, and age—alongside the attributes of potential resettlement locations, GeoMatch predicts where refugees are most likely to find employment. Initial studies demonstrated that using GeoMatch could increase employment rates by approximately 40% in the U.S. and 75% in Switzerland. ​

Dylan Walsh, 2023, HAI Stanford